1. 概述

Semaphore(信号)  并不存在真正的许可 只是维护一个计数器, 通常用来限定进入一些资源的线程数

 accquire()  方法获取许可 成功则计数器值-1 没有则阻塞直到一个可用的许可(即计数器>0)

release() 方法 潜在的释放了申请人(通过给计数器值+1)

2. 示例一(单独测试信号量增减  availabelPermits对超出数量的线程的阻塞)

package com.rocky.semaphore;

import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.Semaphore; public class SemaphoreTest { public static void main(String[] args) {
Semaphore semaphore = new Semaphore(5, false);//no fair 并行最大为5,阻塞后来的
ExecutorService service = Executors.newCachedThreadPool();
for(int i=0; i<10; i++){
service.execute(new Worker(semaphore));
}
service.shutdown();
} } class Worker implements Runnable{ private Semaphore semaphore;
Worker(Semaphore semaphore){
this.semaphore = semaphore;
}
@Override
public void run() {
try {
semaphore.acquire();//获取许可
try {
System.out.println(Thread.currentThread().getName()+" accessing...");
Thread.sleep((long) (Math.random()*3000));
} finally{
semaphore.release();//释放许可
System.out.println(Thread.currentThread().getName()+" leaving...");
}
} catch (InterruptedException e) {
e.printStackTrace();
}
}}

说明:   默认是非公平的 进来一个线程获取许可, 则state减1,直到值为0 以下是源码片段

        final int nonfairTryAcquireShared(int acquires) {//acquires值为1
for (;;) {
int available = getState();//当前state值
int remaining = available - acquires;
if (remaining < 0 ||
compareAndSetState(available, remaining))//remaining>0 则CAS修改state 成功获取许可
return remaining;
}
}   //remainimg<0 返回后,执行下面方法
private void doAcquireSharedInterruptibly(int arg)
throws InterruptedException {
final Node node = addWaiter(Node.SHARED);//创建共享型节点加入等待队列 队列为空则仿制头结点并建立联系
try {
for (;;) {
final Node p = node.predecessor();
if (p == head) {
int r = tryAcquireShared(arg);//再次尝试获取许可
if (r >= 0) {//成功获取许可
setHeadAndPropagate(node, r);
p.next = null; // help GC
return;
}
}
if (shouldParkAfterFailedAcquire(p, node) &&
parkAndCheckInterrupt())
break;
}
} catch (RuntimeException ex) {
cancelAcquire(node);
throw ex;
}
// Arrive here only if interrupted
cancelAcquire(node);
throw new InterruptedException();
}

3. 示例二(一个生产者与一个消费者 两组信号量 此消彼长 为0阻塞)

package com.rocky.semaphore;

import java.util.ArrayList;
import java.util.List;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.Semaphore; public class ProducerCustomRealizeWithSemaphore { public static void main(String[] args) {
SemaphoreBuffer semaphoreBuffer = new SemaphoreBuffer();
Producer producer = new Producer(semaphoreBuffer);
Customer customer = new Customer(semaphoreBuffer);
ExecutorService service = Executors.newCachedThreadPool();
service.execute(customer);
service.execute(producer);
service.shutdown();
} } class SemaphoreBuffer{
List<Integer> list = new ArrayList<Integer>();
Semaphore producerSemaphore = new Semaphore(1);// 允许并行的线程数为1
Semaphore customerSemaphore = new Semaphore(0);// 0即state的初始值 则一开始消费就阻塞了 见上例说明中remaining<0 public void put(int num){
try {
producerSemaphore.acquire();
try {
list.add(num);
} finally{
customerSemaphore.release();
}
} catch (InterruptedException e) {
e.printStackTrace();
}
} public int get(){
try {
customerSemaphore.acquire();
try{
return list.remove(0);
}finally{
producerSemaphore.release();
}
} catch (InterruptedException e) {
e.printStackTrace();
}
return 0;
} } class Customer implements Runnable{
private SemaphoreBuffer buffer;
Customer(SemaphoreBuffer buffer){
this.buffer = buffer;
}
@Override
public void run() {
while(!Thread.interrupted()){
int num = buffer.get();
System.out.println("Customer get the num "+num);
try {
Thread.sleep(2000);
} catch (InterruptedException e) {
e.printStackTrace();
}
} }
} class Producer implements Runnable{
private SemaphoreBuffer buffer;
Producer(SemaphoreBuffer buffer){
this.buffer = buffer;
}
int c =0;
@Override
public void run() { while(!Thread.interrupted()){
buffer.put(c);
System.out.println("Producer put the num "+c);
c++;
try {
Thread.sleep(2000);
} catch (InterruptedException e) {
e.printStackTrace();
}
}
} }

4. 示例三(多个生产者与多个消费者 运用阻塞队列  线程安全 )

package com.rocky.semaphore;

import java.util.concurrent.BlockingQueue;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.LinkedBlockingDeque;
import java.util.concurrent.LinkedBlockingQueue;
import java.util.concurrent.Semaphore;
import java.util.concurrent.atomic.AtomicInteger; public class ProducerCustomRealizeWithSemaphoreLinkedBlockingQueue { static AtomicInteger c = new AtomicInteger(1);
public static void main(String[] args) {
ExecutorService service = Executors.newCachedThreadPool();
CakeStand stand = new CakeStand();
service.execute(new CakeProducer(stand, "producer1", c));
service.execute(new CakeProducer(stand, "producer2", c));
service.execute(new CakeProducer(stand, "producer3", c));
service.execute(new CakeCustomer(stand, "customer1"));
service.execute(new CakeCustomer(stand, "customer2"));
}
} class Cake{
private String name;
Cake(String name){
this.name = name;
}
public String toString(){
return name;
}
} class CakeStand{
BlockingQueue<Cake> queue = new LinkedBlockingQueue<Cake>(15);
Semaphore notFull = new Semaphore(10);//生产信号量
Semaphore notEmpty = new Semaphore(0);//消费信号量
public void put(Cake cake){
try {
notFull.acquire();
try{
queue.put(cake);
}finally{
notEmpty.release();
}
} catch (InterruptedException e) {
e.printStackTrace();
} } public Cake take(){
try {
notEmpty.acquire();
try{
Cake cake = queue.take();
return cake;
}finally{
notFull.release();
}
} catch (InterruptedException e) {
e.printStackTrace(); }
return null;
}
} class CakeProducer implements Runnable{ private CakeStand stand;
private String name;
private AtomicInteger c;
public CakeProducer(CakeStand stand, String name, AtomicInteger c) {
this.stand = stand;
this.name = name;
this.c = c;
}
@Override
public void run() {
while(!Thread.interrupted()){
String str = "cake-"+c.getAndIncrement();
System.out.println("生产:"+name+"-"+str);
stand.put(new Cake(str));
try {
Thread.sleep(1500);
} catch (InterruptedException e) {
e.printStackTrace();
}
} }
}
class CakeCustomer implements Runnable{ private CakeStand stand;
private String name;
public CakeCustomer(CakeStand stand, String name){
this.stand = stand;
this.name = name; } @Override
public void run() {
while(!Thread.interrupted()){
Cake cake = stand.take();
System.err.println("消费:"+name+"-"+cake.toString());
try {
Thread.sleep(1000);
} catch (InterruptedException e) {
e.printStackTrace();
}
}
}
}

说明1): 传统的消费者 生产者使用wait/notify模式等待和相互唤醒, Semaphore通过信号量的值控制运行(>0)和阻塞(<=0),

两组信号量可以使两组角色彼此唤醒,使用阻塞队列可以确保线程安全。

2) 可以额外创建一个信号量Semaphore mutex = new Semaphore(1); 在获取本组信号量之后再获取metex信号量可以实现互斥锁效果

java.util.concurrent.Semaphore 使用的更多相关文章

  1. 161207、高并发:java.util.concurrent.Semaphore实现字符串池及其常用方法介绍

    实现字符串池: StrPool.java import java.util.ArrayList; import java.util.List; import java.util.concurrent. ...

  2. 聊聊高并发(二十五)解析java.util.concurrent各个组件(七) 理解Semaphore

    前几篇分析了一下AQS的原理和实现.这篇拿Semaphore信号量做样例看看AQS实际是怎样使用的. Semaphore表示了一种能够同一时候有多个线程进入临界区的同步器,它维护了一个状态表示可用的票 ...

  3. [转载] java多线程学习-java.util.concurrent详解(二)Semaphore/FutureTask/Exchanger

    转载自http://janeky.iteye.com/blog/770393 ------------------------------------------------------------- ...

  4. Java 并发工具包 java.util.concurrent 用户指南

    1. java.util.concurrent - Java 并发工具包 Java 5 添加了一个新的包到 Java 平台,java.util.concurrent 包.这个包包含有一系列能够让 Ja ...

  5. Java并发编程-并发工具包(java.util.concurrent)使用指南(全)

    1. java.util.concurrent - Java 并发工具包 Java 5 添加了一个新的包到 Java 平台,java.util.concurrent 包.这个包包含有一系列能够让 Ja ...

  6. java.util.concurrent包API学习笔记

    newFixedThreadPool 创建一个固定大小的线程池. shutdown():用于关闭启动线程,如果不调用该语句,jvm不会关闭. awaitTermination():用于等待子线程结束, ...

  7. java.util.concurrent 多线程框架

    http://daoger.iteye.com/blog/142485 JDK5中的一个亮点就是将Doug Lea的并发库引入到Java标准库中.Doug Lea确实是一个牛人,能教书,能出书,能编码 ...

  8. Java_并发工具包 java.util.concurrent 用户指南(转)

    译序 本指南根据 Jakob Jenkov 最新博客翻译,请随时关注博客更新:http://tutorials.jenkov.com/java-util-concurrent/index.html.本 ...

  9. 013-并发编程-java.util.concurrent.locks之-AbstractQueuedSynchronizer-用于构建锁和同步容器的框架、独占锁与共享锁的获取与释放

    一.概述 AbstractQueuedSynchronizer (简称AQS),位于java.util.concurrent.locks.AbstractQueuedSynchronizer包下, A ...

随机推荐

  1. MINIUI监听处理"drawcell"事件

    http://www.blogjava.net/grid/archive/2012/11/02/390693.html     jQuery MiniUI 开发教程 表格控件 表格:自定义单元格(三) ...

  2. 最新的PHP trait使用方法详解

    说通俗点,PHP中使用trait关键字是为了解决一个类既想集成基类的属性和方法,又想拥有别的基类的方法,而trait一般情况下是和use搭配使用的. 具体案例,查看php中文网这篇文章.http:// ...

  3. 对drf序列化器的理解

    序列化: 将对象的状态信息转换为可以存储或传输的形式的过程.(百度定义) 对应到drf中,序列化即把模型对象转换为字典形式, 再返回给前端,主要用于输出 反序列化: 把其他格式转化为程序中的格式. 对 ...

  4. ListView-电影列表

    import React, { Component } from 'react';import { Platform, StyleSheet, Text, View, ListView } from ...

  5. 使用Session监听器实现统计在线人数

    使用Session监听器实现统计在线人数 1.工作目录结构 包含监听器类和jsp页面 2.session监听器 首先利用session监听器来实现对访问网站时的session计数,当有session创 ...

  6. storm(5)-分布式单词计数例子

    例子需求: spout:向后端发送{"sentence":"my dog has fleas"}.一般要连数据源,此处简化写死了. 语句分割bolt(Split ...

  7. kafka监控服务搭建

    wget https://github.com/Morningstar/kafka-offset-monitor/releases/download/0.4.1/KafkaOffsetMonitor- ...

  8. Vue触发input选取文件点击事件

    CSS .upload-btn-box { margin-bottom: 10px; button { margin-right: 10px; } input[type=file] { display ...

  9. Protrator 官网和下载工具稍慢 , 但能使用. Angular CLI 内置 官方推荐 TS的 demo 不明显 , 而且依赖无法安装

    npm uninstall -g protractor  && cnpm install -g protractor && protractor --version 复 ...

  10. (转)haproxy启动故障:Starting proxy:cannot bind socke

    haproxy启动时提示失败: [ALERT] 164/110030 (11606) : Starting proxy linuxyw.com: cannot bind socket 这个问题,其实就 ...